2019년 3월 비트코인 가격 예측 (검증)

예측 기간 : 2월 22일 ~ 3월 20일

# library
library(data.table)
library(plotly)

# read data 
data_w_yhat = fread("C:/Users/user/Documents/BITCOIN/output/backtest_20190222.csv")
data_w_y = fread("C:/Users/user/Documents/BITCOIN/output/realdata_20190320.csv")

# date 
data_w_yhat[,ymd:=as.numeric(gsub("-","",substr(V1,1,10)))]
data_w_yhat[,time:=gsub(" ","",substr(V1,11,19))]
data_w_yhat = data_w_yhat[time=="22:00:00",]
data_w_y[,ymd:=as.numeric(gsub("-","",substr(date,1,10)))]
data_w_y[,time:=gsub(" ","",substr(date,11,19))]
data_w_y = data_w_y[time=="22:00:00",]

# cut range
data_w_yhat = data_w_yhat[ymd>=20190222 & ymd<20190320,list(date=V1,Yhat=daily_avg)]
data_w_y = data_w_y[ymd>=20190222 & ymd<20190320,list(date,Y=close)]

# merge
data = merge(data_w_y,data_w_yhat,by="date",all.y=T)
data[,date:=substr(date,1,10)]

# plot
plot_ly(data, x = ~date) %>%
  add_trace(y = ~Y, type = 'scatter', mode = 'lines', line = list(color = 'blue', width = 1), name="실제값") %>%
  add_trace(y = ~Yhat, type = 'scatter', mode = 'lines', line = list(color = 'red', width = 1, dash='dot'), name="예측값") %>% 
  layout(xaxis=list(fixedrange=F, title="date"),yaxis=list(fixedrange=F, title="price"),autosize = F, width = 900) %>% 
  config(displayModeBar = F)